Online dynamic risk calculator for early detection of stroke

Kemal N. Siregar, Hendy Risdianto Wijaya, Eko Supriyanto, Maheza Irna Mohamad Salim, Tris Eryando, Wan Nor Syuhada

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

Stroke is one of the most common killers worldwide. The death of the most stroke patients is caused by late treatment. In order to prevent the late treatment, an online dynamic stroke risk calculator has been developed. The predictor considers three risk factor groups which are biomolecule related factors, life style related factors and physiological measure related factors. Age, gender, family history and own disease history are factors related to biomolecule structure. Food, water, and air intakes as well as environment, physical and mental activities are the main components of life style. Measurement of blood pressure, heart rate, blood glucose, body mass index, blood lipid are done to provide information on physiological measure related factors. In order to improve the accuracy of risk calculator, clinical assessment is recommended to be conducted if the risk factor is above human average risk at specific age. This includes to assess sign, symptom, stroke biomarker, as well as stroke related anatomy and pathophysiology imaging. All factors and their levels are stored in the cloud database. Online application to enter data and visualize results is connected to the cloud database. Rule-based algorithm and machine learning are applied to calculate the risk of getting stroke. More than 12,000 retrospective global data are stored in the database. The database and rule are dynamically updated by the new online data input to improve accuracy of prediction through learning process. The risk calculation (rule-based algorithm) has been compared with other algorithms on machine learning to prove the system model. The system has been also validated using 120 healthy data and 25 stroke patient data. Test result shows that the system produces more than 95% accuracy and can be a better dynamic stroke risk predictor that can be applied to machine learning. This system applies to early detection of stroke.

Original languageEnglish
Title of host publication3rd Biomedical Engineering''s Recent Progress in Biomaterials, Drugs Development, and Medical Devices
Subtitle of host publicationProceedings of the International Symposium of Biomedical Engineering, ISBE 2018
EditorsPraswasti P.D.K. Wulan, Misri Gozan, Sotya Astutiningsih, Ghiska Ramahdita, Radon Dhelika, Prasetyanugraheni Kreshanti
PublisherAmerican Institute of Physics Inc.
ISBN (Electronic)9780735418226
DOIs
Publication statusPublished - 9 Apr 2019
Event3rd International Symposium of Biomedical Engineering''s Recent Progress in Biomaterials, Drugs Development, and Medical Devices, ISBE 2018 - Jakarta, Indonesia
Duration: 6 Aug 20188 Aug 2018

Publication series

NameAIP Conference Proceedings
Volume2092
ISSN (Print)0094-243X
ISSN (Electronic)1551-7616

Conference

Conference3rd International Symposium of Biomedical Engineering''s Recent Progress in Biomaterials, Drugs Development, and Medical Devices, ISBE 2018
Country/TerritoryIndonesia
CityJakarta
Period6/08/188/08/18

Keywords

  • Dynamic System
  • Early Detection
  • Machine Learning
  • Risk Calculator
  • Risk Factor
  • Stroke

Fingerprint

Dive into the research topics of 'Online dynamic risk calculator for early detection of stroke'. Together they form a unique fingerprint.

Cite this